Automated Monitoring and LoRaWAN Control Mechanism for Swiftlet Bird House

Farmers in Malaysia are moving into edible swiftlet bird nest cultivation simply because it's another potential growing industry with high income in South East Asia. The biggest challenge in this industry is how to simulate the swiftlet's habitat inside the bird house to attract more birds. The suitability of the environment especially inside these houses are largely unknown due to very little research effort in this new area. This paper presents how wireless sensor networks (WSN) such as LoRa and video analytics techniques are utilized to simulate a conducive environment for the birds to stay in this artificial bird house. The wireless sensors are installed in the bird houses to collect humidity, temperature and oxygen data. With the combination of above data plus the number of birds entering and leaving the bird nest will be used to simulate an optimum environment inside the bird house. The bird counting algorithms developed throughout this project shows an accuracy of 92.5%.